Implementation of Exploratory Data Analysis (EDA) in Predicting Customer Churn Decision Levels Using Naive Bayes Algorithm

Abstract

This study aims to explore the use of Exploratory Data Analysis (EDA) for predicting customer churn decisions at service provider companies. Using the Naive Bayes algorithm within the RapidMiner environment, the research analyzes customer behavior patterns to determine key influencing factors. The process involves structured data preprocessing, visualization, and model evaluation through accuracy, precision, recall, and F1-score metrics. The results show that EDA coupled with the Naive Bayes model provides valuable insights and reaches an accuracy of over 82%, making it a reliable decision-support tool for customer retention strategies. This study contributes to data-driven approaches in customer relationship management.